TPS65987D: TPS65987D internal power MOSFET damaged at 4.8A

Alan Chow

Part Number: TPS65987D
Other Parts Discussed in Thread: TPS55288, TPS65988

Dear sir,

I have encountered the problem about the TPS65987D’s internal power MOSFET broken during loading out around 4.8A, 20V. My design is mainly referenced to the TPS65987EVM and TPS55288 is used as buck boost converter which connects to the TPS65987D by I2C. In addition, only PP_HV2 power path is enabled for DFP configuration. Please see the schematics below. 5V, 9V, 15V and 20V PDOs are broadcasted once the USB loading plugs in. The input voltage to the TPS55288 is 25.2V.

During debug process, E-mark cable was used and the PDO 20V 5A was selected. At the beginning, the USB loading started to load 3A first and the loading current slowly increase. Until reached to 4.8A, output voltage drops and the loading current became zero.

I found that the internal MOSFET of TPS65987D has broken out. Besides, the CC1 and CC2 couldn’t communicate to the loading any more. However, its I2C communication to the TPS55288 still works with. The TPS65987D’s LDO_3V3, LDO_1V8 also alive.

Why did the internal MOSFET burn out at 4.8A? Should the hardware design need to change to duo power mode configuration for 5A? The TPS55288 part seems to work fine after this issue.

Why CC1 and CC2 couldn’t communicate after that? Actually, except for the communication, CC pins could still pull down to 1.6V when it connected to the USB loading after this issue.

Is the problem related to the E-mark cable? However, 5V supply has already supplied to the pp_cable pin for VCONN in my design. Or the layout problem?

Please help to give me some hints.

Thank you!

Best regards!

Alan

over 4 years ago

0 Chuck Branch over 4 years ago

TI__Mastermind 32921 points

Alan,

It is not clear from your description if the power path of the TPS65987D has been damaged or if it simply turned off which can have several causes.

First: If you repeat your test, does the part reconnect and negotiate your target contract?

If it does, then have you checked you settings for the TPS55288 and ensured that it is configured to support 5A in it's internal current limit. If the Buck Boost current limit is set too low, then the behavior that you described makes sense. You will need to update the current limit of the buck boost to match your PDO contract.

If neither of these is true, then let me know and we can debug further.

0 Alan Chow over 4 years ago in reply to Chuck Branch

Prodigy 140 points

Dear Chuck,

Thank you for the reply! I have tried to repeat the the test with the same PCBA after the issue. In my setting, once the USB loading attach, TPS65987D sends out I2C signal to TPS55288 to output 5V. Unfortunately I saw that TPS55288 could output 5V correctly at its output side after receiving I2C signal from TPS65987D. However there is no 5V output at the VBUS2 of TPS65987D. In addition, the USB loading didn't receive any PDO from the CC pin. My TPS65987D project setting is attached below and please see anything wrong first. Thank you for your quick support!!

TPS65987DDH_20V_5A(8-Sept-2021).pjt

Best regards!

Alan

0 Chuck Branch over 4 years ago in reply to Alan Chow

TI__Mastermind 32921 points

Alan,

I have checked your project and I do not see any issues with it. Are you able to confirm that the I2C events that control the TPS55288 and set it to 5V occur before you connect the type C UFP? This is the first step I would address on debug.

Regards,

Chuck

0 Alan Chow over 4 years ago in reply to Chuck Branch

Prodigy 140 points

Chuck,

Let me do the test again to reconfirm it for your question as soon as possible. Thank you!

Best regards!

Alan

0 Chuck Branch over 4 years ago in reply to Alan Chow

TI__Mastermind 32921 points

I will hold this open until I hear back from you.

0 Alan Chow over 4 years ago in reply to Chuck Branch

Prodigy 140 points

Chuck,

Sorry about that I can't re-check the I2C with the damaged PCBA. I remember that, when TPS65987D booted up, it passed the 5V setting to the TPS55288 but the TPS55288 output was not enabled. Until the TPS65987D detects cable attached (Note: The E-mark cable was used in the debug process), TPS65987D requested TPS55288 to enable its 5V output.

Best regards!

Alan

0 Chuck Branch over 4 years ago in reply to Alan Chow

TI__Mastermind 32921 points

Alan,

The TPS65988 needs to be enabled at POR at 5V. This should resolve your issue with connection when you get a functional board to work with again.

0 Alan Chow over 4 years ago in reply to Chuck Branch

Prodigy 140 points

Chuck,

Sorry for my late reply. I will set the output 5V at POR for debug testing. However, the new PCBA will arrive after a few days. Once I have the new PCBA and tested, I will tell you the result. Thank you for your help!

best regards!

Alan

0 Chuck Branch over 4 years ago in reply to Alan Chow

TI__Mastermind 32921 points

Glad to help

Interface

Interface forum

TPS65987D: TPS65987D internal power MOSFET damaged at 4.8A